How can I easily backup & restore bugtracker of project on Github? Repository itself is obvious, wiki is also a git repository - but what about bugtracker?
I want to fully backup bugtracker (tickets, attached imaged, labels, comments, images attached in comments, dates of creation and closing of tickets etc) with ability to recreate it on github account.
Saving it as editable format would be a good bonus.
I am aware about Github API and that I can make tool like this. But I would prefer to use (optionally improve) existing one.
Existing ones mainly export the issue text, not always their attachments.
github-issues-export
github-to-bitbucket-issues-migration
Those services could be extended with the issue API to export more.
A more complete service is backhub.co, which exports a bit more (but isn't as easily expendable). It is free for public repos.
Update August 2018: vorburger adds in the comments a direct link example of how to use the GitHub Issues API to obtain all comments from a particular issue:
https://api.github.com/repos/JDKMissionControl/jmc/issues/1/comments
Related
Azure DevOps includes a Wiki feature for all projects. The wiki content is stored as a git repository, with a .md markdown file for every page. You can clone the repository and look at how it is organised. All pages in the wiki allow users to submit comments - clearly these must be stored somewhere, but where? They do not appear in the wiki repository.
Unfortunately, we haven't sort it into any history and it only relevant to Wiki page until now. This comment feature is being implement but haven't supported from UI.
As of today, what we provide is view the comments content from UI, or get them from API:
Get https://dev.azure.com/{org name}/{project name}/_apis/wiki/wikis/{wiki name}/pages/{page id}/comments
In fact, the backend data structure has been built, but we haven't provide the corresponding UI to public. So, with above api, you can not only get the comments you want, but also record the modification history of the comments.
Personally, it sounds a good idea about putting comments into source control since it not only records some important conversation about current wiki, also has representation about its development process.
Much recommend you can share your idea here, it is the official forum that our Product Group reviewed.
You can share the suggestion ticket link here, I'm sure I would be the first voter on that. Also, there will have other SO users who view this ticket support that.
Is there any way to disable viewers to download my files from github?
I want to show my work, but I am afraid anybody can steal my code.
You can't, Github actually means you want to share something with the community, or you are showcasing your work in the open source community. Unfortunately you cannot restrict the users from downloading your content from your Repository.
Alternatively what you can do Just make your Repository private. But then in this case you are not able to make you work available to view for audience.
More you can read here :-
https://help.github.com/en/github/building-a-strong-community/limiting-interactions-in-your-repository
Disabling zip/tarballs on GitHub at the moment seems to be impossible, but it would be useful for those using git submodules, which are not included in the automatic created archives, where the repository owner could replace them with a continuous integration job with something like git-archive-all.
Is there a way I can make Jira ticket id's on Github link to Jira? (much the same way as how it happens on Bitbucket).
Maybe there is a plugin or something I could use ?
EDIT:
I specifically want ticket id's in Github to link through to Jira, this is not described in any of the resources below:
How do I connect github to JIRA?
https://confluence.atlassian.com/display/BITBUCKET/Linking+Bitbucket+and+GitHub+accounts+to+JIRA
https://help.github.com/articles/integrating-jira-with-your-projects/
Update 2022:
Dusan Plavak mentions in the comments that, as of today (Q1 2022), you can also import / automate creation of autolinks via github api.
You can also use gh api:
gh api repos/octocat/hello-world/autolinks \
-f -q 'key_prefix:key_prefix url_template:url_template'
And since July 2022, you also have autolinks with alphanumeric IDs (for instance for a Trello card URL: https://trello.com/c/3eZr2Bxw).
2019: No need for Chrome plugin, Safari hack or scripts.
From October 2019, GitHub can automatically transform TICKET references to Jira links, for example.
See "Save time linking resources with autolink references" from Lars Schneider.
Now you can set up an autolink reference and GitHub will automatically create links (to external systems) for you.
How it works
If you use GitHub with external services (like Jira), you might be familiar with shorthand references (like TICKET-123) to point to resources in those external systems.
Starting today, GitHub can automatically transform shorthand references into clickable links for GitHub Pro, Team, and Enterprise plans.
So not the regular github.com.
To enable this feature, a repository admin needs to register the reference prefix (such as TICKET-) in the settings.
Afterwards, references of that type are detected in issues, pull requests, comments, or commit messages, and turned into links.
This speeds up navigation between GitHub and external systems.
See documentation.
Note that the ticket prefix is not included in the hyperlink despite being included in the anchor text. The URL preview in the setting is accurate:
The animation in the feature announcement, replicated above, shows this limitation but not clearly.
This means that you have to either include the project key in the target URL as well or prefix all project keys with a another key and use that as the autolink prefix.
You cannot even use Git's trailers to reference tickets and use autolinks, because trailers are key-value pairs separated by : (colon space) and an autolink prefix that ends with a space is invalid.
I decided to make a Chrome extension for this ! This extension works on all websites, not just Jira ! :D
Install it here:
https://chrome.google.com/webstore/detail/jira-hotlinker/lbifpcpomdegljfpfhgfcjdabbeallhk
Source code:
https://github.com/helmus/Jira-Hot-Linker
If you mean the commit message for commit e0d9e32 shows TEST-1234 Fixed this really bad bug and you want TEST-1234 to link to http://www.your-jira.com/browse/TEST-1234 then no it does not look like this is possible.
You could look to do a Chrome/Firefox browser plugin that accomplishes the behavior you desire. (Scans the webpage looking in certain sections for JIRA Key's and then updates them to links)
I wrote a site specific Safari hack. It requires you to install it yourself, but that gives you the opportunity to read the code (it's short!) first. https://github.com/unicode-org/icu-jira-safari
it fetches elements of class commit, gh-header-title, and js-issue-row with getElementsByClassName()
then, for anything such as ICU-1234 (our Jira project id), it creates a link to the appropriate URL on the Jira server.
I wanted to include Jira issue status as well, but Atlassian Cloud CORS settings prevent that.
Here is an example if the linkification (ICU-10464 turns into a link on the right hand side).
Someone made exactly what I wanted coming from gitlab.
https://chrome.google.com/webstore/detail/linkify-jira-issues/ekbbnaokafbanjgmcbllligemhiclbcb
This updates the text in my github PRs to convert them in clickable links.
I made a super simple Chrome extension for this. It simply adds a link to the relevant JIRA ticket on your Github pull request page in a non-obtrusive way (See screenshot below)
Install it here: https://chrome.google.com/webstore/detail/github-to-jira/dhnagjladcclacjnccpnibdmkljidnhl?hl=en-GB
Source code here: https://github.com/mfalade/github-to-jira
While Github Autolink is an awesome solution, this is easier to use for the following reasons.
You don't need to provide a description with ticket number for this to work. It automatically derives ticket number from your branch name.
You don't have to configure anything on the Github admin page. Perfect for users who don't have admin privileges
You can work on multiple projects or teams more easily. Supports multiple ticket prefixes.
This extension works for all project management tools you might want to link to
I'd like to see a graph of the total size of a repo over time.
(this is useful as it can help to see if poeple are accidentally checking in stuff they shouldn't, like binaries etc.)
The standard github graphs show lots of interesting stuff, but not this simple metric so far as I can see.
Any ideas?
Have you considered giving a try to the GitHub Archive project?
As stated on the home page
GitHub Archive is a project to record the public GitHub timeline, archive it, and make it easily accessible for further analysis.
GitHub Archive provides downloadable json datasets and also makes its content available via Google BigQuery for easy online analysis (see this document for more information on this topic and sample queries).
Note: GitHub timeline data is available starting February 12, 2011.
Update:
Provided your repository is private, you may have to fallback to the GitHub API and consider the size member. In order to graph, you'd have to trigger a regularly call to the API and locally store the result.
syntax: /repos/:user/:repo [GET]
example: https://api.github.com/repos/libgit2/libgit2
[[Please note that Github has changed a lot since this question was first asked; instead of "download page" read "new release".]]
I generate PDF documentation as part of my projects and I'd like them to stay in sync with my Git repository (it's not always possible for people to build their own since they often use proprietary fonts).
However, it's not really "correct" to add the PDFs to the repository since it's a derived file; furthermore, doing this adds significantly to the size of the commits and the size of the repository overall.
Is it possible to programmatically send files to the GitHub download page? (I know that tagged commits are automatically added there with git push --tags but I don't know where this is documented. I suppose I could do something fancy by adding a separate branch only containing the PDFs themselves — as done by the GitHub user pages — but I'm a bit rusty on using Git this way.)
Github API v3 supports this feature.
GitHub also provides a maven plugin based on the java API that is part of the Eclipse Mylyn connector for GitHub.
There is a ruby gem called github_api.
The other answer talks about net-github-upload which is available for perl and ruby.
check out for net-github-upload which is available
for perl: http://github.com/typester/net-github-upload-perl
and ruby: http://github.com/Constellation/ruby-net-github-upload
With that you can write a small script to upload and update your PDF easily. To sad there's no easy way provided by github guys themselves..
cheer!
The GitHub blog post announcing that this feature has been disabled: https://github.com/blog/1302-goodbye-uploads
I take it that by "GitHub download page", the owner means a repo–more specifically a branch– that can be downloaded via the "download" button.
If you want to add a file to a repo using the API, you will have to become familiar with the process described here: https://developer.github.com/v3/git/
It's not the easiest process in the world, but mastering it will force you to understand the concepts of blobs, trees, commits and references, amongst others.
You can't just "send a file" to a repo because you're working with Git, and Git has some "internal expectations" that you just can't ignore (it's impossible to think of GitHub as some sort of host that you can ftp). Explaining the flow required to create a file in a GitHub repo is certainly beyond the scope of the original question, but to provide a clear answer: no, it's not possible to programmatically upload a file on GitHub, but yes it is possible to programmatically push a file on GitHub".
There's a PHP library named GitHubTreePHP that lets you automate the process (Disclaimer: I wrote it).